Table querying

Abstract

The subject invention relates to a system and / or methodology that facilitate converting structured data (e.g., tabular information) into natural language, thus making it available to question answering systems and search engines. More specifically, the system and method employ the vast quantities of natural language on the particular storage system, database, network, or server that a user wishes to search. For example, the system can utilize natural-language based information located on the Web to assist in the task of converting structured data into natural language. In particular, the system can find at least one sentence or sentence fragment on the Web that refers to at least one tuple (e.g., row) in a given table. Following, those sentences can be generalized across all or substantially all tuples in the table. The resulting body of text can be used by the question answering system to answer user queries.

Description

TECHNICAL FIELD [0001] The subject invention relates generally to searching for information and in particular to searching data maintained in a tabular or other structured form indirectly by associating such data with related data in a searchable language form. BACKGROUND OF THE INVENTION [0002] The amount of available information or data maintained such as on the World Wide Web (“Web”) is vast and almost limitless. However, not all of the information may be easily accessed by a user. For example, current search engines or question answering systems are unable to take advantage of the large quantity of information contained in tabular format including tables, charts, fact lists, etc. This is largely because they depend on information that is encoded as sentences in natural language. As a result, conventional question answering systems severely restrict the amount of information retrievable by the user. SUMMARY OF THE INVENTION [0003] The following presents a simplified summary of th...

AI Technical Summary

Benefits of technology

[0004] The subject invention relates to a system and / or methodology that facilitate converting structured data (e.g., tabular information) into natural language, thus making it available to question answering systems as well as search engines. More specifically, the system and method employ the vast quantities of natural language on the particular storage system, network, or server that a user wishes to search. For ease of explanation, imagine that the user performs a search for information on the Web, for example. The system can utilize the information stored on the Web to assist in the task of converting structured data into natural language based on existing natural language found on the Web. In particular, the system can find at least one sentence or sentence fragment on the Web that refers to some of the tuples (e.g., rows) in a given table. Following, those sentences can be generalized across all or substantially all tuples in the table. The resulting body of text can be used by the question answering system to answer user queries. Essentially, the subject invention can leverage the vastness of the text on the Web to determine the relationships between items maintained in tabular form as well as the sentence syntax with which to state those relations. Once stated in natural language, any language based information retrieval or question answering system can make use of the data.

[0005] According to one aspect of the invention, the structured data can be maintained in a non-sentence format and can be converted to a natural or grammar-based language by locating existing natural or searchable language corresponding to at least a portion of the table and then determining the relationships between the portion of the table and the existing searchable language. The existing searchable language may be found on one or more web pages, for instance, and portions of the existing searchable language may match or be similar to at least a portion of data in the table. Portions that match as well as any other relevant text that precedes or follows can be extracted and employed to generalize or convert substantially all the data in the table into a searchable language format. In practice, the structured data can also be converted to non-grammatical language formats as well such as templates (e.g., $person born $date), for example. Moreover, the structured data can be converted to any language or format that facilitates querying.

[0007] In addition, the AI component can be trained explicitly or implicitly to determine which tables, such as on the Web, are more useful to the user than others. Therefore, the less useful tables are less likely to undergo the conversion process

Problems solved by technology

The amount of available information or data maintained such as on the World Wide Web (“Web”) is vast and almost limitless.

However, not all of the information may be easily accessed by a user.

For example, current search engines or question answering systems are unable to take advantage of the large quantity of information contained in tabular format including tables, charts, fact lists, etc.

As a result, conventional question answering systems severely restrict the amount of information retrievable by the user.

Images